Microstructure – The microstructure of the studied Hybrid steels is shown in the optical micrographs in Fig. 1. Both microstructures are apparent from the martensitic lath structures. The tempered microstructure contains finely-dispersed carbides and nickel aluminides. These precipitates can only be observed with electron microscopy or atom probe tomography 18. The as-rolled Hybrid 55 steel has a hardness of 35 HRC, which is increased by tempering to 55 HRC.
Electrochemical passivation behaviour in acidic solutions – The potentiodynamic polarization behaviour of the Hybrid steels in de-aerated sulfuric acid is summarized in Fig. 4a,b. The polarization behavior of the stainless-steel specimens, the IF steel and the pure metals (nickel, chromium, and aluminium) are shown for comparison purposes only. The polarization behavior of stainless steel is well-known in the literature; the discussion is focused on understanding the polarization behaviour of Hybrid steel. In de-aerated sulfuric acid (pH 0.5), the as-rolled and tempered Hybrid steel show an active-to-passive transition peak and a broad passive region until transpassive breakdown. The critical passivation current density of the as-rolled Hybrid steel is higher than the stainless steels but lower than IF steel, indicating an improved passive behaviour compared to IF steel. However, in the pH 3 acidic solution, the IF steel and tempered Hybrid steel showed no passivation but spontaneous active dissolution. However, the as-rolled Hybrid steel underwent passivation in the same solution after a wide active dissolution during anodic polarization beyond the corrosion potential. The critical passivation current densities for the stainless steel and Hybrid steel ranged from 10− 6 to 10− 4 A/cm². The as-rolled Hybrid steel showed an extended passive range compared to all measured steels, including 304 stainless steel, at high anodic polarization between approximately 900 mV and the water reduction potential for both measured pH solutions (see Fig. 4). The polarization curves showed several peaks as a function of anodic polarization. The activation peak beyond the corrosion potential comprises at least three specific elemental dissolution activities. Electrochemical data read from the polarization curves are summarized in Table 3.
Table 3
Electrochemical data extracted from the polarization curves presented in Fig. 4.
Condition
|
Material
|
Passivation Potential (V)
|
Critical Passivation Current (A/cm2)
|
Passivation Current (A/cm2)
|
Transpassive Potential (V)
|
pH 0.5
H2SO4
|
As rolled Hybrid steel
|
0.15
|
1·10− 1
|
3.7·10− 5
|
1.5
|
Tempered Hybrid steel
|
0.6
|
6·10− 1
|
1.7·10− 4
|
1.49
|
420 stainless steel
|
-0.31
|
6·10− 3
|
9.4·10− 6
|
1.518
|
304 stainless steel
|
-0.222
|
1·10− 5
|
2.7·10− 6
|
0.91
|
Interstitial-free steel
|
0
|
1·10− 1
|
1.1·10− 4
|
1.15
|
pH 3
H2SO4
|
Non-Tempered Hybrid steel
|
0.1
|
9·10− 5
|
2.0·10− 6
|
1.34
|
Tempered Hybrid 55
|
Spontaneously active
|
420 stainless steel
|
-0.30
|
1·10− 5
|
2.7·10− 6
|
1.744
|
304 stainless steel
|
-0.34
|
6·10− 6
|
2.3·10− 6
|
1.346
|
Interstitial-free steel
|
Spontaneously active
|
Surface chemistry analysis – To further understand the polarization behaviour of Hybrid steel, we performed HAXPES analysis using synchrotron x-ray radiation. We polarized both Hybrid steels to various anodic polarization levels as indicated in Figure S 1 (see supplementary file) and Table 2 to understand why the as-rolled condition showed passivation behaviour and the reason for the minor activation peaks in the polarization curve. The oxidation behaviour of the tempered steel was similar to the low-carbon steel (IF steel). Therefore, only one polarization state was chosen. Specimen no. 1 and no. 6 are non-polarized conditions forming native surface oxide/oxyhydroxide. The survey HAXPES spectra are given in Fig. 9. Several core-level peaks of the alloying elements can be seen. Some Auger peaks are also labelled. The Cr 2p, Fe 2p, Ni 2p, Mo 3p, and O 1s core level peaks were scanned in high resolution. The Mo 3p core level has been chosen instead of Mo 3d due to better intensity-to-background signal ratios. The Al 1s core level peaks are not shown, but these were also scanned in high resolution. These gave far more intensity yield than the Al 2p signals and were, therefore, chosen for the analysis. The Hybrid steel contains minor silicon, which was disregarded in the HAXPES analysis. Samples 6 and 7 showed additional peaks, which were identified as carbides. Tempering produces mainly Cr-rich and V-rich carbides and nickel aluminides (NiAl), which were not distinguishable from the Ni 2p core level spectra. Transition metal carbides can be several tens of eV’s away from the carbon core level peak.
The high-resolution HAXPES analysis results for the main alloying elements are shown in Fig. 10 (iron), Fig. 11 (chromium), Fig. 12 (nickel), Fig. 13 (molybdenum), and Fig. 14 (aluminium). The core level peak ratios of the metallic and oxidation states change as a function of anodic polarization and microstructure (Fig. 15). The native oxide of Hybrid steel is composed of Fe, Cr, Ni, Mo, and Al oxides/oxyhydroxides. The as-rolled steel contained considerably more aluminium and nickel oxide than the tempered steel. The HAXPES results of sample 3 (as-rolled) and sample 7 (tempered) show that the surface chemistry evolved after polarization to 500 mV vs Ag/AgCl (see Table 2 and Figure S 1). Thus, the aluminium and chromium oxide species are mainly responsible for the passivation of the as-rolled Hybrid steel. The HAXPES fitting results for each peak model defining the composition of the surface film of Hybrid 55 steel are listed in detail in Table S 1. At this point, we want to restate our rigour and scrutiny in HAXPES measurement and analysis. It is a common weak practice in the literature to report background-subtracted and fitted XPS data only without showing the unmanipulated data. In such a way, the reader has no clue about the actual data and may not entirely understand the photoemission mechanism. Therefore, we present the original data with the models used for data analysis and allow the reader to ponder without forced manipulation.
Thermochemical modelling for Hybrid steel in acidic solutions – The thermochemical modelling results are given in Fig. 16, Fig. 17, and Fig. 18. Figure 16a shows the calculated phase diagram section for the H2O – O2 – H2SO4 – Steel system (four components). “Steel” is the chemical composition of Hybrid steel but as the sum of aqueous ionic and solid compound products. The diagram shows the formed solid corrosion products and their molalities (y-axis) as a function of added metal concentration (x-axis) at initial pH 0.5 with a fixed oxygen partial pressure of 10− 1 atm. The oxygen partial pressure has a nearly linear relationship (see Fig. 16b) with the redox potential of the steel-electrolyte system. Therefore, 10− 1 atm oxygen partial pressure was deliberately chosen to reflect the polarization curve (at around 1.3 V in Fig. 4) at high anodic potentials (“High Potential”). Figure 16a shows that Fe2O3 forms at lower metal cation concentrations in the electrolyte (the system described above), followed by NiSO4(H2O)7 and then Al2O3(H2O). The formation of the latter two products requires more metal dissolution, hence, more (intense) polarization. So, at high potentials, these products are thermodynamically stable, and at least these contribute to the passivity of Hybrid steel.
It should be noted that the modelling results only represent a local equilibrium that takes a long polarization time and ignores kinetic considerations. The x-axis represents the metal concentration in the local equilibrium, which includes all of the metal-oxide compounds assumed to be formed on the steel electrode and metallic cations and anions in the electrolyte. Figure 16c shows that the interfacial pH change drastically increased with the metal cation concentration. The minor interfacial pH change at low metal ion concentration is due to the formation of Fe2O3. The oxygen needed for iron oxide formation is delivered primarily from water molecules, resulting in dissociation and hence a rise in pH. The interfacial pH increases sharply when nickel sulphate and aluminium oxide are formed. Furthermore, the sharp rise in pH is associated with the amount of Fe2O3 and the formation of NiSO4(H2O)7 and Al2O3(H2O) due to water dissociation and crystalline water bound to each product molecule. No chromium oxide exists when the redox potential is high enough to dissolve chromia.
The corrosion products formed at low redox potential are depicted in Fig. 17. “Low” in this context represents the passive state of the polarization curve of Hybrid steel (at around 500 mV in Fig. 4). The modelling results show that Fe2O3, NiSO4(H2O)7 and Al2O3(H2O) form first during polarization, followed by the production of Cr2O3 and (NiO)(Fe2O3). It should be noted that the latter compound could have also been calculated as NiO. So, the oxide product shall be understood as nickel oxide. We disagree with the classic understanding that surface oxides are layered or stratified. However, there is a chemical gradient along and across the surface film in all geometrical dimensions. All oxides build together one complex film entity on the surface. Compared to high potentials, chromia and nickel oxide are stable at low potentials. Therefore, the formation of chromia increases the interfacial pH, like iron, nickel, and aluminium products. A steep rise in pH occurs when (NiO)(Fe2O3) forms, shifting the interfacial pH by more than two orders of magnitude. These explain the passive behaviour of Hybrid steel in de-aerated pH 0.5 sulfuric acid solution. The steepest rise in pH occurs when (NiO)(Fe2O3), a spinel-type oxide, forms, indicating a significant contribution to passivity.
The computational modelling results are summarized in Fig. 18a, showing the formed corrosion products and their molalities as a function of oxygen partial pressure (fugacity) in pH 0.5 sulfuric acid solution. The x-axis (logarithmic oxygen concentration) can be qualitatively understood as the Eh of a classic Pourbaix diagram and the redox potential of a standard electrochemical polarization curve. Oxygen pressure on the order of 10− 50 – 10− 60 atm is highly deficient and indicates practically no anodic polarization. So, it can be deduced that the spinel oxides FeCr2O3, (NiO)(Fe2O3), and Al2O3(H2O) form the native oxide of Hybrid steel in the acid. The latter two compounds remain stable throughout polarization until the breakdown potential of the Cr(III)/Cr(VI) couple redox potential sets on, which is at about 750 mV in the polarization curve (see Fig. 4). Figure 18b shows that the oxygen fugacity and the Eh have a nearly linear relationship. With increasing oxygen fugacity (≈ electrode potential, Eh), FeCr2O3 transforms to Cr2O3, indicating selective oxidation of iron. As a result, more Fe2O3 is produced, forming the most surface oxide. The anodic dissolution of Cr2O3 and NiO occurs at similar potentials. The amount of NiSO4(H2O)7 increases indicating a conversion of NiO to NiSO4(H2O)7. The dissolution of Cr2O3 reduces the interfacial pH by more than two orders of magnitude. Thus, producing Cr6+ (aq) ions causes severe acidification due to water hydrolysis.
The thermodynamic computation in 0.1 M chloride solution predicts Fe2O3, NiO, Cr2O3 and Al2O3 as the native surface oxide passive film of Hybrid steel (Fig. 19). Like sulfuric acid’s reaction, FeCr2O3 transforms to Cr2O3, indicating environment-independent behaviour. We performed the calculations with and without CO2, a typical electrolyte compound due to its ubiquitous presence in the air and its immediate absorption by the aqueous solution. The solution’s initial pH was about 6.5, measured before the experiment. FactSage showed a pH of 6.4 when we added 10− 5 atm CO2 fugacity, and after the simulation, the system’s pH (the interfacial pH) became minor alkaline. Hence, the natural presence of CO2 is insignificant and, therefore, negligible.
The surface is protective until Ni2+ (aq) and Cr6+ (aq) are formed before the breakdown (dissolution) of NiO and Cr2O3. The degradation of the passive film occurs when the molality of the solid oxide compounds reduces in magnitude, especially with the dissolution of Cr2O3. It should be noted that the removal of Cr2O3 is triggered by the excessive formation of Cr6+ (aq), which reduces the interfacial pH by more than two orders of magnitude (Fig. 19). Cr6+ ions form several complex molecule ions and have a strong hydrolysis effect. We have calculated the corrosion products with their molalities as a function of oxygen fugacity for pure iron, chromium, nickel, and aluminium and have seen no pH change except for chromium. The calculations indicate that the pH reduction reasons the dissolution of Cr2O3. The reduction of the interfacial pH also dissolves NiO but favours Fe2O3 production. Al2O3 is only slightly affected by the pH and, therefore, takes a more decisive role in passivity since it remains on the surface.
Electrochemical pitting tests in chloride-containing solutions – The results of the polarization behavior in the near-neutral chloride solutions are presented in Figure 5. The as-rolled Hybrid steel show similar polarization behavior to the grade 420 martensitic stainless steel. The corrosion current densities at the corrosion potential for each tested steel are on the order of sub-micro to microamps. All tested steels show a steep rise in anodic current density after a few hundred millivolts beyond the corrosion potential. A steep climb in anodic current density in chloride-bearing solutions is a sign of the breakdown of the surface oxide that imparts resistance to anodic oxidation, hence, pitting corrosion. Thus, the polarization curves indicate an existence of a breakdown potential and hence passivity. Since no active behavior was observed, the data suggests spontaneous passivity for the Hybrid steels and the martensitic stainless steel. The 420-steel showed minor current fluctuations close to the breakdown potential, indicating metastable pitting events. In contrast, the Hybrid steels led to no metastable pit activities in both tested chloride solutions. The passive current densities of the Hybrid steel were on similar levels as the martensitic stainless steel. The latter showed similar breakdown potential to the as-rolled Hybrid steel. The polarization behavior of the IF steel showed an immediate rise in anodic current density, indicating spontaneously active corrosion. The polarization curves were reproducible, with a maximum deviation of 50 mV.
Corrosion morphology – After the polarization tests in the 0.01 M chloride-containing solution, the corrosion morphology was examined with an optical microscope (Fig. 6). Like the 420 martensitic stainless steel, the Hybrid steel showed several corrosion pits in the exposed area, indicating the existence of a protective passive surface oxide and its local breakdown on several sites. The pits on the Hybrid steel were hemispherical and dish-shaped, whereas lacy-cover pits with a perforated layer formed on the martensitic stainless steel. Localized corrosion on Hybrid steel was also observed in the more aggressive chloride-concentrated solution (Fig. 7 and Fig. 8). However, the corrosive attack manifested itself in the form of bands, which were more severe in the tempered microstructure. The bands were about 50–100 microns in width and elongated along the rolling direction. The banded corrosion morphology is due to a preferential attack on an elemental deficient region caused by solidification and hot deformation during alloy production, known as macro-segregation 18.
Macro-segregation is a common metallurgical issue in steel production. The segregation can be noticed in the optical micrograph shown in Fig. 1. The compositional variation in the microstructure has been reported in earlier communication 18. Alloying elements, mainly nickel and chromium, vary perpendicular to the rolling direction. Differences in composition lead to the formation of micro-galvanic cells, which becomes more apparent when the corrosion severity is high. This effect was not seen in the diluted chloride solution. Micro-galvanic corrosion occurs when the corroding region is active while the remaining part is passive, driven by the oxidation power strength of the electrolyte (and polarization). Such a phenomenon is typical, for example, for duplex stainless steel 24. In earlier work, it has been shown that the pitting corrosion of duplex stainless steel in diluted chloride solution is pitting, which selectively occurs in the ferrite phase 24,25. The corrosion concentrates more on the entire ferrite phase when the chloride concentration is increased due to more pit nuclei forming next to each other in the ferrite phase. It has been shown that selective corrosion is pitting corrosion with very high density. In this work, the situation is practically the same. Segregated sites showing more noble alloy composition behave as net cathodes whereas elemental depleted sites become net anodes. Interestingly, despite the high polarization above the breakdown potential, the major corrosion occurred on the segregation bands.